zivicmilos/llm-performance
收藏Hugging Face2023-10-03 更新2024-03-04 收录
下载链接：
https://hf-mirror.com/datasets/zivicmilos/llm-performance
下载链接
链接失效反馈官方服务：
资源简介：
---
# For reference on model card metadata, see the spec: https://github.com/huggingface/hub-docs/blob/main/datasetcard.md?plain=1
# Doc / guide: https://huggingface.co/docs/hub/datasets-cards
{}
---
# Dataset Card for LLM Performance
### Dataset Summary
This table presents a comprehensive comparative analysis of a few popular LLMs, such as Falcon, LLama 2, and Mistral, highlighting both the quality of their outputs and the corresponding inference times. We finetuned the Falcon model with the full Alpaca dataset of 52k datapoints and with randomly sampled 5k datapoints and then compared them with base and instruct versions of Falcon, LLama 2 and Mistral. All models are with 7B parameters and in int4 representation.
提供机构：
zivicmilos
原始信息汇总
数据集卡片：LLM性能
数据集概述
该数据集提供了一个全面的比较分析，涵盖了几种流行的LLM模型，如Falcon、LLama 2和Mistral，重点展示了它们输出质量及相应的推理时间。我们使用完整的Alpaca数据集（52k数据点）和随机抽样的5k数据点对Falcon模型进行了微调，并与Falcon、LLama 2和Mistral的基础版本和指令版本进行了比较。所有模型均具有7B参数，并以int4表示。
